TL;DR: A poorly briefed sample request is the single fastest way to burn 3–4 weeks on a requote cycle that never needed to happen — get your colour tolerance sign-off documented before anything goes to press.
TL;DR: In our intake process, briefs missing a stated ΔE 2000 acceptance limit cause an average of 2.3 additional approval rounds before production sign-off.
What Your Brief Actually Needs to Say About Colour Before We Can Quote #
Most of the delay in sample and quotation cycles for colour-managed packaging comes from the brief itself, not from the factory. When a brand partner submits a request without a defined spectrophotometric tolerance, our colour team has to make assumptions — and assumptions become revisions.
Before requesting samples for any packaging that carries brand colour, you need three things documented in your brief:
1. A reference standard. This means a physical Pantone swatch number, an existing approved press sheet with a stated CIELAB value (L, a, b*), or a digital file with embedded ICC profile (minimum: ISO Coated v2 or GRACoL 2013 profile for CMYK builds). A logo PNG does not count. A rendered PDF without a profile attached does not count.
2. A stated ΔE tolerance. ISO 12647-2:2013 defines acceptable ΔE 2000 deviation for process colours on coated paper at ≤3.0 for primaries and ≤5.0 for secondary colours. For premium brand packaging, we work to tighter internal targets: ΔE 2000 ≤ 1.5 on spot brand colours and ≤ 2.0 on critical secondary tones. If you do not specify a limit, our team will apply our default P-CAL/04 colour acceptance gate, which uses ΔE 2000 ≤ 2.5 — but that may or may not match what your brand team expects.
3. Substrate and finish confirmation. Colour reads differently on gloss-coated, uncoated, soft-touch laminated, or kraft surfaces. A spot colour targeted on coated white stock will shift by as much as ΔE 5–8 when printed on natural kraft without adjustment. Provide the substrate first; we calibrate our spectrophotometer measurement conditions (D50 illuminant, 2° observer, M1 or M2 measurement mode per ISO 13655:2017) to match the actual intended surface.
| Brief Element | Minimum Acceptable Input | What We Use If Missing |
|---|---|---|
| Colour reference | Pantone code or CIELAB Lab* values | sRGB screen render — high requote risk |
| Tolerance | ΔE 2000 numeric limit | Our P-CAL/04 default: ΔE 2000 ≤ 2.5 |
| Substrate | Named grade + surface finish | Our standard 350 gsm coated art stock |
| ICC/profile | Named characterisation dataset | ISO Coated v2 assumption |
| Print process | Offset, digital, flexo | Offset assumed, may change cost tier |
When all five elements are present at intake, our quoting turnaround for colour-managed rigid or folding carton packaging runs 3–5 working days. When they are not, that extends to 8–12 working days as we chase clarifications.
Where Sample Requests Break Down: Three Failure Patterns We See Repeatedly #
The first and most common failure is submitting artwork at screen resolution. We receive PDF or AI files prepared at 72 dpi or RGB colour mode for digital display. When these go to our RIP (Raster Image Processor), the output at 150 lpi screen ruling looks visibly degraded — fine type under 6pt will fill in, and any tonal gradient will show banding. Artwork for offset print must be submitted at minimum 300 dpi at final output size, CMYK or spot colour channels, with a 3 mm bleed on all cut edges. If the file does not meet this, we flag it under our intake step AW-CHECK/01 and hold the sample until a corrected file arrives. This adds 2–5 working days.
The second pattern is requesting a printed proof before confirming structure. We regularly receive proof requests before the structural die-line has been finalised. A printed proof produced to an unconfirmed die-line has a meaningful chance of being scrapped when dimensions change — the brand pays for a proof that cannot be used as an approval reference. Our process requires a signed-off white sample (unprinted structural sample) before any colour proof goes on press. The white sample stage typically takes 7–10 working days for custom tooling on rigid boxes and 4–6 working days for folding carton dielines we already carry in our standard library of approximately 340 die templates.
The third pattern is requesting a production sample (GP3 — gold production sample) without first completing a colour profile match on our specific press-substrate combination. Colour profiles built for a press at a previous supplier do not transfer. Our spectrophotometers run on an X-Rite i1Pro 3 platform, calibrated twice daily against a ceramic white reference tile traceable to ASTM E308 illuminant standards. The first time we run your colour on our press, we build a press characterisation dataset from a 1617-patch ECI2002 target. That dataset becomes your account’s colour fingerprint with us. Trying to skip this step and go directly to production sample approval means any out-of-tolerance result has no validated baseline to reference — and the colour argument becomes circular.
Does the Sample Type Affect the Quotation Price? #
Directly: yes, though not by much. White samples (structural only, no print) are typically provided at cost of materials plus tooling where new dies are required. A white sample in a standard size requires no new tooling and costs less than $30 USD in most cases. A custom die adds a one-time tooling charge of $80–$250 USD depending on complexity.
Printed colour proofs on our digital proofing press cost more per unit but require no minimum. Production samples run on the offset press have a practical minimum of around 50–100 sheets to achieve press stability — below that, the first 20–30 sheets are stabilising and the colour data from them is not representative of production. Quoting based on production samples below that floor gives misleading cost-per-unit figures.
When comparing quotes from multiple suppliers, request that each quote specifies: the press type and screen ruling used, the substrate grade and caliper (in mm, not just “350 gsm coated”), and whether the quoted price includes a colour profile build or assumes an existing profile. A quote that omits these variables is not directly comparable to one that includes them, regardless of the unit price.
Specification Notes for Brand Partners #
When you brief us on packaging that requires colour sign-off, the single most useful thing you can send is an approved physical sample from your current production — even if it is from a different supplier. We scan it on the i1Pro 3, build a target CIELAB value set, and work backwards to formulate inks and press settings to match. That eliminates interpretive guesswork on both sides.
The brief gap that causes the most sample iterations is misalignment between the digital reference used for design and the physical substrate used for production. A colour that looks correct in Photoshop on an uncalibrated monitor, approved via email screenshot, and then printed on an uncoated natural board will almost always require at least two correction rounds. Send us your intended substrate before we quote. If you do not have a preference, tell us the product category and we will recommend from our standard stock grades.
Our sampling sequence runs: white sample (4–10 working days depending on tooling), colour proof on digital press (3–5 working days after approved white sample and confirmed artwork), and production sample off-press (5–8 working days after proof approval). Total typical sequence: 12–23 working days. Rush options exist for stages 1 and 3 at additional cost, not stage 2 — digital proof speed is already near its practical floor.
Frequently Asked Questions #
What file format should I send for sample artwork?
Send print-ready PDF with all fonts embedded, CMYK or named spot colours (no RGB), minimum 300 dpi at final print size, with 3 mm bleed on all cut edges. Native AI or InDesign files are acceptable if you include all linked assets and fonts, but PDF/X-4 is our preferred receive format because it preserves transparency handling and colour profiles without ambiguity.
Can I request a digital proof instead of a press proof to save time?
Yes, and for most packaging categories it is the right first step. A digital proof on our calibrated Epson proofing system, run to ISO 12647-7:2016 proofing standard, accurately predicts offset press output within ΔE 2000 ≤ 2.0 for most CMYK builds. Where it diverges is on metallic inks, fluorescents, and any colour simulated by the proofing system rather than laid down as a physical match — in those cases, a press proof on substrate is the only reliable approval document.
How do I know if my colour tolerance requirement is realistic for offset printing?
It depends on the colour. Process CMYK builds on coated stock can hold ΔE 2000 ≤ 2.0 run-to-run with G7 calibrated press conditions, per the IDEAlliance G7 Master specification. Spot colours mixed from Pantone system inks hold tighter — ΔE 2000 ≤ 1.5 is achievable across a full run on stable stock. The challenge is the first press sheet versus the 10,000th: colour drift over a long run is real, and our inline spectrophotometric monitoring flags deviation beyond our P-CAL/04 gate (±ΔE 2000 1.8 drift threshold) and triggers an operator adjustment before the run continues. Tolerances tighter than ΔE 2000 1.0 on process colour are outside what offset press physics reliably deliver without prohibitive waste rates.
What quantity should I request for a production sample, and does it count toward my order?
We typically pull 10–20 units from a production sample run. They count toward the production run if the run meets our minimum order quantities — typically 500 units for rigid boxes and 1,000 units for folding cartons — and the sample is approved without requiring a corrective re-run. If corrections are needed, the sample units are scrapped and not credited.
How do I compare quotes fairly when different suppliers quote on different substrates?
Insist on identical substrate specifications in each quote: paper grade (art card, SBS, duplex, etc.), caliper in millimetres, surface finish (gloss/matt coated, uncoated), and FSC certification status if required. A 350 gsm gloss-coated SBS board from one supplier and a 350 gsm uncoated duplex from another are not comparable products — the price difference does not reflect supplier efficiency, it reflects different materials. FSC Chain of Custody certification (FSC-STD-40-004) status also affects cost and lead time and must be declared at quote stage if your brand policy requires it.
What is the biggest mistake brands make when evaluating their first production sample?
Approving or rejecting it under non-standard lighting. Metamerism — where two colours match under one light source but diverge under another — is a real effect, particularly for colours in the red-orange range and for any sample containing optical brighteners. Always evaluate production samples under a calibrated D50 light source (5000K, CRI ≥ 90). Phone cameras and office fluorescent lighting will give you a different read than retail shelf conditions. If you cannot evaluate under D50, flag that to your colour contact and request a cross-illuminant metamerism check against D65 and A (tungsten) before sign-off.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
Submitted a brief for a limited-edition run — 18,000 units, metallic silver spot on soft-touch laminate — and we didn’t specify a ΔE tolerance because the brand team assumed the converter would “match the approved sample.” They hit ΔE 2000 of 3.8 against our reference and considered it a pass under their house standard. Took three requote cycles and six weeks to sort out that nobody had put a number on paper. The article’s point about P-CAL/04 defaults is exactly the gap that burned us.
The ΔE 2000 ≤ 1.5 target on spot brand colours is realistic for a single-substrate run, but we’ve found that tolerance basically falls apart when you’re producing the same SKU across both litho-printed folding carton and a flexo-printed shrink sleeve simultaneously — we had to negotiate a ΔE 2000 ≤ 2.8 cross-process tolerance with our licensee in Shenzhen last year just to keep the shelf set looking coherent. The brief still needs the tolerance, but it needs to specify which process it applies to or you’re just creating a different argument later.
We ran into this on a cosmetics secondary pack last year — same substrate, same Pantone 485 C spec, but half the run came back on uncoated board because the brief didn’t distinguish between inner tray and outer carton. The uncoated stock pushed ΔE 2000 to 4.1 against the approved coated press sheet. Two requote rounds and six weeks lost.
We started attaching a physical press approval sheet with recorded Lab* values directly to every brief after we had a coconut wax tin label come back visually acceptable but technically outside our internal gate — turned out the converter had been working from an sRGB export the whole time.
The ICC profile requirement catches people out more than the ΔE spec does, in my experience — we had a folding carton brief come in with a GRACoL 2013 tagged PDF for the outer, but the inner tray artwork was supplied as an untagged file, and our colour team applied P-CAL/04 defaults to both. Took us until the second physical sample round to realise the inner was being proofed against a different baseline entirely, which explained a visible shift on the brand’s primary blue even though both pieces technically passed the ΔE 2000 ≤ 2.5 gate.
The part about assumptions becoming revisions is where we’ve lost the most time — our Q3 2023 refresh cycle for a retail electronics accessory line had three briefs come back for requote purely because the substrate confirmation was missing a finish spec, not the colour reference itself, and each loop was 8–11 working days by the time the factory queue reshuffled.
The P-CAL/04 default catching people out is real — we had a brief go to our Shenzhen supplier last spring for a protein bar secondary carton where we hadn’t locked a tolerance because the colour (a deep forest green, mixed spot) had always “just matched” across four previous runs on the same press. They applied their house default, which was actually ΔE 2000 ≤ 3.0, and the samples passed their QC but failed our brand gate at ≤ 1.8. Two rounds of press proofs and six weeks later we finally got it resolved by just writing the number into the brief, which we absolutely should have done in 2019 when we first onboarded them.